A Wari Tapestry Textile in a Tiwanaku Tomb from the Osmore Valley, Moquegua, Peru

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Textile Society of America Symposium Proceedings Textile Society of America

9-2012

A Wari Tapestry Textile in a Tiwanaku Tomb from the Osmore Valley, Moquegua, Peru

Amy Oakland California State University - East Bay, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/tsaconf

Oakland, Amy, "A Wari Tapestry Textile in a Tiwanaku Tomb from the Osmore Valley, Moquegua, Peru" (2012). Textile Society of America Symposium Proceedings. 719. https://digitalcommons.unl.edu/tsaconf/719

This Article is brought to you for free and open access by the Textile Society of America at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Textile Society of America Symposium Proceedings by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln.

A Wari Tapestry Textile in a Tiwanaku Tomb from the Osmore Valley, Moquegua, Peru Amy Oakland [email protected]

For the Textile Society of America 2012 panel on Royal Patronage and Textile Collections my contribution concerns patronage and textiles from the Andean region of South America. This is also a paper that discusses the most elite tapestry textiles created in the period known as the Andean Middle Horizon, between the 7th and 11th centuries when the two cultures Wari and Tiwanaku co-existed. My own discoveries this summer of Wari and Tiwanaku textiles together in the same tomb present a new view of these cultures, one that identifies interaction, rather than the usual separation between these two important pre-Inca groups.

As scholars like John Murra (1962), John Rowe (1980), Ann Rowe (1978), and Elena Phipps (2004) have discussed, textiles were considered the most precious material to the Inca and single weft- interlocked tapestry was the most prized of all Inca textiles. Inca men’s tapestry tunics were worn as upper-body coverings or shirts that draped from the shoulders to the knees with an opening at the neck and on the sides for the arms. Ann Rowe (1978:8-13) discovered an unfinished Inca tapestry tunic that demonstrated that the textile itself was woven as a single web, finished on all four sides with the neck slot created by placing discontinuous warps over a scaffold. Considering the topic of this TSA session, Inca interlocked tapestries were undoubtedly royal Andean garments. No Inca tapestries have been discovered in the highland Inca capital at Cuzco where seasonal rainfall normally destroys all perishable objects. However remarkable Inca tapestry has been recovered in frozen highland ritual burials and in coastal burial contexts.

The fine Andean tapestry tunics that survive from earlier Wari and Tiwanaku were also woven in single- interlocked tapestry, folded at the shoulders with the warp oriented horizontally. In technique and general size and shape (approx. 1m square) they are so similar to Inca men’s tunics that these Wari and Tiwanaku men’s tapestry tunics must have been valued in a similar manner as elite garments connected to centralized imperial authority. Like the Inca, the capitals of Wari and Tiwanaku were centered in the Andean highlands where few textiles survive, but desert conditions along much of the Pacific coast and inland river valleys have preserved cloth of both cultures. Wari spread from its southern Peruvian capital in Ayacucho to the north in the highlands and along the Pacific south, central and north coast. Tiwanku’s influence outside of its core region in the Bolivian altiplano near Lake Titicaca has been detected west and south in the southern altiplano and adjoining valleys and coastal desert. Recent studies have also discovered that Wari spread south into the Majes and Siguas valleys of the far south coast and that the two groups met in the Osmore Valley near the modern town of Moquegua (Goldstein 2005; Tung 2007; Williams 2001). Here in southern Peru is the only place where Wari and Tiwanaku sites have been discovered in the same valley.

Wari Tapestry Construction

Structural evidence suggests that Inca, Wari and Tiwanaku men’s tunics were woven on a wide rectangular frame loom and warped on the short direction so that the warp will run sideways or horizontally when the tunic is worn. But instead of using a scaffold to create the neck slot, Wari weavers made tunics using two separate individually woven webs each cut from the loom and sewn up the center and sides of the tunic leaving a slot for the neck and armholes (Bird and Skinner 1974). On one loomend of a Wari tunic the heading cord was removed leaving open warp loops that are chained. The chained edges are usually placed together at the center and connected by a dense figure-8 stitch that creates a vertical seam down the center of the Wari tunic. The other cut edge is interlaced diagonally creating a solid finish that is hidden under embroidered side selvedges. Most Wari tapestries are warped with two- ply (Z2S) white cotton. But as Bird and Skinner (1974) noticed Wari tapestry warps could be cotton, camelid fiber (probably alpaca) or bichrome warps of both camelid fiber and cotton. More than 300 Wari tunics have survived and are today displayed in major world museums, but few were excavated with any archaeological provenance. Angeles and Pozzi-Escot (2000) discuss Wari men’s tapestry tunics and other Wari textiles recently excavated on Peru’s central coast at Huaca Malena. Most of the 21 Wari tapestry tunics discussed by Prumers (1990; 2000) uncovered at “El Castillo” on Peru’s north coast included Z2S cotton warps. But he found that some Wari tapestries were woven with all camelid fiber and one even used S-spun warps and another was warped with Z3S replied yarns (Prumers 1990:183). Highland and southern Andean spinners typically use the drop spindle that produces a Z-spun yarn and these are plied together (Z2S). North coast spinners use a spindle with a small whorl held horizontally that produces S-spun yarns off the spindle tip and these were often paired and not replied. As more Wari textiles are analyzed in archaeologically known contexts it may be possible to discover original weaving centers. However, as for the Inca, the earlier Wari probably exacted textile tribute for redistribution and the find site will not necessarily identify the original weaving center.

Figure 1, left. Wari Tapestry Tunic, Huaca Cao, El Brujo. Photo by A. Oakland. Figure 2, right. Detail, Wari Tapestry Tunic, Huaca Cao. Photo by A. Oakland.

The Huaca Cao Wari tapestry (figures 1 and 2) excavated at El Brujo in the Chicama Valley on Peru’s north coast, far from the Wari center identifies all of the most common Wari tapestry features: two-web

2 construction, white cotton Z2S warps chained along one loomend and connected at the tunic’s center covered with fig-8 stitches over the vertical central seam (Rodman and Fernandez 2000, 2005).

Tiwanaku Tapestry Tunics

Very few Tiwanaku tapestries have survived and most of these are so fragmentary that they are not on public display. All known Tiwanaku tapestry tunics excavated in archaeologically recorded collections were woven like Inca tapestry tunics as a single finished four-selvedge web with the neck slot woven over a scaffold that holds discontinuous warps. Tiwanaku weavers used only camelid fiber in elite men’s single-interlocked tapestry tunics. Tiwanaku tapestry (Figure 3-4) uncovered in San Pedro de Atacama remains among the finest and best-preserved tapestry known from the region of Tiwanaku influence (Oakland 1986).

Figure 3, right. Tiwanaku Tapestry Tunic 5382, Museo Padre Le Paige, San Pedro de Atacama. Figure 4, left. Detail Tiwanaku Tapestry Tunic 5382. Photo by A. Oakland.

Gustavo Le Paige excavated the six-banded Tiwanaku tapestry 5382 in a burial in Coyo Oriente and wrote “painted textile” in his notebook mistaking the exceptional tapestry with 12 warps and 64 wefts per cm (Oakland 1986). Other Coyo Tiwanaku tapestries were woven with paired Z2S warps with as many as 22 warps and 91 wefts per cm. A rare complete Tiwanaku two-banded tapestry tunic with a bright yellow ground was discovered in a Bolivian cave burial at Pulacayo with 13 warps/ 46 wefts per cm (Aguero 2007:90-91). The overall design of most of these Wari and Tiwanaku tapestries presents vertical tapestry bands between solid vertical stripes with narrow tapestry bands along the side selvedges. Repeating images include animal-headed staff-bearing figures, but geometric steps, frets, and images that repeat mirrored and divided in four parts are known to tunic patterning in both cultures. Wari and Tiwanaku tapestries woven without vertical bands are also known.

Tiwanaku and Wari Tapestry in Omo 16D-tomb 15, Moquegua

In July and August 2012 I was able to examine a portion of the remarkable textile collections excavated by Paul Goldstein in the central Osmore Valley near Moquegua in southern Peru. For its proximity to the Tiwanaku capital one would expect this to be a region with Tiwanaku affiliation. Goldstein (2005) has determined that Tiwanaku colonists settled in several places in the Valley farming the valley floor and building houses and a Tiwanaku temple at Omo. He excavated extensive domestic and funerary contexts during all periods of Tiwanaku occupation in Moquegua discussed as Omo, Chen-Chen, and Tumilaca style (Goldstein 2005:134). At the same time in the same valley Wari constructed, some would say a royal palace on the top of the prominent natural plateau Cerro Baul (Williams 2001). The Wari center overlooked Tiwanaku settlers who farmed the valley below; a situation that is archaeologically very clear but culturally perplexing in the ridged separation of Wari and Tiwanaku sites in the valley and in the distinct cultural material that these two entities seem to have maintained for at least four centuries (Goldstein 2005: 165). No textiles have been uncovered in the wetter environment at Cerro Baul just 20kms north of Moquegua. But in the Tiwanaku colonies Goldstein has excavated the greatest amount of Tiwanaku textiles yet known from both domestic and burial contexts.

The Moquegua excavations provide the perfect location to review Tiwanaku perishable culture since no textiles or basketry have been preserved at the highland capital. My analysis focused on Tiwanaku domestic textiles excavated in a large refuse deposit among Chen-Chen period houses at Rio Muerto and in earlier pre-Tiwanaku Huaracane textiles, but I report here on a discovery I made when looking at Omo-style textiles from the exceptionally large tomb M16D, tomb15. Goldstein (2005: 261-264) described the tomb as one of the most elite Tiwanaku burials yet excavated, but the tomb had been destroyed sometime in antiquity. The remains of a man and two children were discovered disarticulated and contents strewn in a wide area around the immense tomb, visible on the surface by a ring of fallen stones covering over 7m in diameter. A fragmentary group of weft-interlocked tapestry textiles had remained preserved together apparently dating to the period of the tomb AD 635-890 (Goldstein 2005:263).

Figure 5. Provincial Tiwanaku Tapestry Omo M16-5140c, Museo Contisuyo. Photo by A. Oakland.

4 Goldstein (2005:263) stated that “three distinct tapestry shirts were found as a mass sandwiched together”, but I discovered that this analysis might not tell the entire story. Listed all under the same number 5140 there were three different Tiwanaku tapestries in the tomb, but actually four tapestries in all and one of them was a Wari tapestry formed of cotton warps with a chained warp selvedge (see Figs. 9-10 below).

One Tiwanaku tapestry fragment (5140c, figure 5) was courser than the rest (5-6 warps and 19-25 wefts per cm.) with thick, tightly twisted brown camelid fiber (Z2S) warp and red, green, gold, black, blue, and dark green camelid-fiber tapestry (Z2S) wefts. The fragment (27 x 19.5cm) and larger parts not yet conserved identify that this large textile was completely patterned, but the design remains unidentified.

Evidence of the loomend and embroidered side selvedge suggests that the tapestry was originally a tunic. Prominent features include wefts of a great variety of thicknesses giving this a “provincial” quality like Tiwanaku tunics from Ilo and San Pedro de Atacama (Minkes 2008; Oakland 1986). This thicker tunic was found directly placed against an extremely fine tapestry (5140b).

This second tapestry (5140b, Figure 6) with all brown camelid-fiber (Z2S) warps and red, green-blue, dark blue, gold, cream and dark brown (Z2S) weft colors appears more like other known classic Tiwanaku tapestry examples. The weft yarns are extremely fine and evenly spun with 14 warps and 50 wefts per cm., twice as fine as the first tunic (5140c).

Figure 6. Tiwanaku Tapestry Omo M16-5140b, Museo Contisuyo. Photo by A. Oakland.

Because of the small size of the existing fragment (13.5 x 11.5 cm.) the overall design is ultimately undecipherable, but the large scale, and the rounded outlines of images that are only parts of much larger figures identify that this tapestry would not fit into a standard two or four-banded tunic. The tapestry may have been patterned with wide vertical bands, but the evidence of the solid color between tapestry bands does not exist and it is equally possible that this would have been a classic-style Tiwanaku

5 tapestry with an overall design. Perhaps new Tiwanaku tapestry image types will be identified as more textiles are excavated in controlled contexts as here in Moquegua.

A third small (10x11.2cm) fragment of a very fine Tiwanaku tapestry tunic (5140a, Figure 7-8) was perhaps mistaken for a portion of the tapestry textile 5140b. But the warps of this tapestry 5140a are predominately blue with one brown warp placed between every 4, 6, or 8 blue warps.

Figure 7, left. Tiwanaku Tapestry Tunic fragment 5140a, Museo Contisuyo. Photo by A. Oakland. Figure 8, right. Detail, Tiwanaku Tapestry 5140a, Museo Contisuyo. Photo by A. Oakland.

The warps in this fragment are also slightly larger (12 warps per cm.) but the (Z2S) wefts were spun so finely and evenly that at least 70 wefts were counted in one cm. (62-70 wefts per cm.). Weft colors include dark blue, blue, green, deep red, red, light gold, and cream. The image is entirely different with a smaller scale figure present on the side of a vertical solid blue band showing that this fragment was originally part of a six-banded tapestry tunic with a dark blue ground. Even though the preserved image is tiny the parts preserved on both sides of the vertical band suggest that the image repeated running staff-bearing figures (see Tiwanaku image in Figure 4 above). The figure itself is missing, but this type held a human captive out in front instead of a staff. The tassels of the ends of the “sacrificer’s” neck sash are visible in this tiny fragment on the left side of the vertical blue band. On the opposite side of the vertical band a small section of the captive’s distinctive face is visible with the round eye, nose, and hat.

I realized that the fourth textile was a Wari tapestry (5140d, Figure 9 and10) that had been associated with the individuals in the elite Tiwanaku tomb 15 at Omo 16D. Wari tapestry features are present: cream-colored or light brown cotton (Z2S) warps chained together along one loomend partially visible in Figure 10. Some of the warps that were originally chained are now open warp loops showing where the heading cord was removed. The loomend section that remains complete identifies that the final textile was finished with an embroidered edge finish of polychrome cross-knit loop stitch. The row of cross-knit loop stitch continued around at least one side selvedge of this Wari tapestry panel, now incomplete, but larger than 64 x 51cms.

Figure 9, left. Omo 16 Wari Tapestry 5140d, Museo Contisuyo. Photo by A. Oakland. Figure 10, right. Detail Loomend of Omo 16 Wari Tapestry 5140d.

The other loomend (not pictured) is also partially intact with approx. 9 comes. bare warp loops left incomplete as if removed from the loom and placed in the tomb as an unfinished offering. It is also possible that the original wefts in this part are now destroyed. The strong cotton warps remain and these are uncut on this loomend and instead are now curling back around adjacent warps because of tight over twisting. Dyed (Z2S) weft colors include red-brown, dark brown, gold, yellow, blue, blue-black, green, and tan hues.

All Omo 16D-5140 textiles, the provincial Tiwanaku tapestry (5140c), the two fine classic Tiwanaku tapestries (5140a and 5140b), and this Wari tapestry (5140d) were woven with wefts in natural camelid shades and dyed in a variety of colors including red, yellow, green, and blue. Both of the important Andean red dyes Relbunium and Cochineal have been identified in textiles from the Osmore Valley (Wallert and Boytner 1996) and Cochineal is found naturally and is cultivated commercially in the valley today. Yellow shades are produced from a variety of natural sources and yellow dye or dyed yarns are often combined with indigo to produce green. Indigo was undoubtedly the source for the blue color in these Moquegua textiles. Zarro (2007:172) identified indigofera cultivated or tolerated in late Chiribaya agricultural fields in the coastal Osmore Valley and he noted that Indigo was a local plant present in the Osmore Valley today. But no one knows where these tapestries were originally woven or if both Wari and Tiwanaku weavers would have had access to the same dye sources. Blue remains a basic color in Tiwanaku tapestry textiles, but it seems that Wari weavers did not have the same dye knowledge or access to indigo sources. Many Wari tapestries use no blue or green color at all (Stone 1987). The tapestry wefts present in the Omo M16 Wari textile are less brilliantly dyed and notably the blue-colored sections include streaks of brown yarns representative of less control over the original dyeing or the addition of “filler” yarns of a color close-in-shade where there might not be enough to fill an area.

As stated above fiber choices and loom preparation remain distinct between Wari and Tiwanaku. Wari textiles regularly combine cotton and camelid-fiber, especially in decorated textiles and Tiwanaku textiles are overwhelmingly produced with camelid fiber only. Tung (2007:259) describes the Wari village at Beringa in the central Majes Valley with large quantities of spindles, weaving tools, and

7 unprocessed cotton fiber in most domestic areas and cotton and camelid-fiber textiles including Wari tunics recovered from looted burials. These data as well as the construction details of Wari tunics support the idea of Wari workshops and cloth tribute for the state.

In the Moquegua Tiwanaku colonies and in the Tiwanaku influenced areas of the South-Central Andes in general, most people wove and wore clothing made from warp-faced plain weave, probably created on staked looms or frame looms similar to looms used to create the same cloth in the altiplano today. Very few tapestry fragments have been recovered in domestic middens in Moquegua and they are rare in Tiwanaku period tombs, but like the Omo-style tapestry group reported here tapestry is present in exceptional circumstances. And in this M16 tomb Wari and Tiwanaku textiles were recovered together.

Since Goldstein discovered these textiles all “sandwiched” together they are probably one portion of a single mummy bundle, most likely associated with the adult male in this Omo-style burial, the most elaborate Tiwanaku tomb yet excavated in the Moquegua Tiwanaku colonies. The tapestry is not a garment. There is nothing structurally evident in the textile that would suggest that it was anything other than a single, flat, rectangular tapestry panel patterned with a large image. Wari tapestry “hangings” or “banners” excavated along the south Peruvian coast have been identified (Conklin: 1971:15; Menzel 1977: Fig 130) and complete miniature tapestries have been remarkably preserved in the actual Wari site (Rodman and Fernandez 2000: Fig 9A, 9B). A tapestry offering would make sense in an Omo tomb in the Tiwanaku colony. It is reasonable to consider placing a valued textile into the grave of a respected individual, but a Wari textile in a Tiwanku tomb? Wari material is rarely encountered in Tiwanaku contexts in Moquegua (Goldstein 2005:166-170) and no Wari textile has ever been discovered here before the present analysis.

Goldstein (2005:165) notes that although Wari and Tiwanku constructed settlements as close to one another as 20 km, they remained “utterly unlike” one another in almost any way and he considers the infrequent finding of material between the two as a suggestion of “an uneasy truce, if not outright hostilities between the outliers of the Wari and Tiwanaku cultures of the Osmore drainage” (2005: 168). But the identification of a Wari textile placed in an elite Tiwanaku tomb allows a consideration that sometimes Wari and Tiwanaku connected and, at least during the event associated with this elite Tiwanaku man buried in this tomb, the two cultures did find some appreciation of the other. What kind of interaction might this be? Perhaps what Alfonso Ortiz describes for the Navaho and Hopi as a “violence-spotted symbiosis”, raiding one another as well as marrying, but maintaining distinct cultural identity. Toward the topic of today’s seminar, the collections housed in the Museo Contisuyo in Moquegua allow new understanding of the major Middle Horizon cultures that must have interacted in various ways. These valuable collections await much needed patronage to expand facilities and provide conservation for the perishable treasures like the Tiwanaku and Wari textiles I have described today.

8 References Cited

Angeles, Rommel and Denise Pozzi-Escot 2000. “Textiles del Horizonte Medio. La Evidencias de Huaca Malena, Valle de Asia”, In Huari y Tiwanaku: Modelos Vs Evidencias, Boletin de Arqueologia PUCP 4:401-424.

Aquero, Carolina, 2007. “LosTextiles de Pulacayo y Las Relaciones Entre Tiwanaku y San Pedro de Atacama. Boletin del Museo Chileno de Arte Precolombino Volume 12 (1):85-98.

Bird, Junius and Milica Skinner, 1974. “The Technical Features of a Middle Horizon Tapestry Shirt from Peru”, Textile Museum Journal, Volume IV(1): 5-13.

Conklin, William, 1971. “Peruvian Textile Fragment from the Beginning of the Middle Horizon”, Textile Museum Journal, Volume 11I (1):15-24.

Goldstein, Paul, 2005. Andean Diaspora, The Tiwanaku Colonies and the Origins of South American Empire. University Press of Florida.

Minkes, Wynne 2008, “Warp the Loom-Wrap the Dead Trapezoid Shaped Textiles from the Chiribaya Culture, South Peru, AD 900-1375”, Textile Society Symposium Proceedings. Paper 116. http://digitalcomons.unl.edu/tsaconf/116.

Menzel, Dorothy 1977, The Archaeology of Ancient Peru and the Work of Max Uhle. R. H. Lowie Museum of Anthropology, University of California, Berkeley.

Murra, John, 1962.“Cloth and Its Functions in the Inca State”. American Anthropologist Volume 64 (4): 710-728.

Oakland, Amy, 1986 “Tiahuanaco Tapestry Tunics and Mantles from San Pedro de Atacama, Chile”. In The Junius B. Bird Conference on Andean Textiles, 1984, p. 101-121. The Textile Museum, Washington D.C.

Phipps, Elena, 2004 “Cumbi to Tapestry: Collection, Innovation, and Transformation of the Colonial Andean Tapestry Tradition”, pp.72-99.In The Colonial Andes: Tapestries and Silverwork, 1530- 1830. Edited by E. Phipps, J. Hecht, and C. Esteras M. Metropolitan Museum of Art.

Prumers, Heiko , 1990. “Der Fundort ‘El Castillo’ im Huarmeytal, Peru. Ein Beitrag zum Problem des Moche-Huari Textilstils”, Mundus Reihe Alt-Amerikanistik 4, Bonn. 2000, “’El Castillo’ de Huarmey: una plataforma funeria del Horizonte Medio” In Huari y Tiwanaku: Modelos Vs Evidencias, Ed. P. Kaulicke and W. Isbell. Boletin de Arqueologia PUCP 4:289-312.

Rodman, Amy Oakland and Vicki Cassman, 1995 “Andean Tapestry: Structure Informs the Surface”. Art Journal 54(2):33-39.

9 Rodman, Amy Oakland and Arabel Fernandez 2000, “Los Tejidos Huari y Tiwanaku Comparaciones y Contextos”, In Huari y Tiwanaku: Modelos Vs. Evidencias, Primera Parte. Eds. P. Kaulicke and W. Isbell, Boletin de Arqueologia PUCP, 4: 119-130. 2005, “North coast Style after Moche: Clothing and Identity at El Brujo, Chicama Vlley, Peru”. Us and Them: Archaeology and Ethnicity in the Andes, ed. R. Reycraft. Cotzen Institute of Archaeology, University of California, Los Angeles.

Rowe, Ann 1978, “Technical Features of Inca Tapestry Tunics”, Textile Museum Journal, Volume 17: 5-28. 1986, “Textiles from the Nasca Valley at the Time of the Fall of the Huari Empire.” The Junius B. Bird Conference on Andean Textiles, ed. A. Rowe. p. 151-182. The Textile Museum, Washington D.C.

Rowe, John 1980, “Standardization in Inca Tapestry Tunics”, The Junius B. Bird Conference on Pre- Columbian Textiles, Washington, D.C.

Stone, Rebecca 1987, Technique and Form in Huari-style Tapestry Tunics: the Andean Artist, A.D. 500-800. Ph.D. diss., Yale University.

Tung, Tiffiny 2007,“The Village of Beringa at the Periphery of the Wari Empire: A Site Overview and New Radiocarbon Dates”. Andean Past Volume 8: 253-286.

Wallert, Arie and Ran Boytner 1996. “Dyes from the Tumilaca and Chiribaya Cultures, South Coast of Peru”. Journal of Archaeological Science 23, 853-861.

Williams, Ryan 2001, “Cerro Baul: A Wari Center on the Wari Frontier”. Latin American Antiquity 12(1)67-83.

Zaro, Gregory 2007, “Diversity Specialists: Coastal Resources Management an Historical Contingency in the Osmore Desert of South Peru”. Latin American Antiquity 18(2):161-179.